This study aimed at determining the intra- and inter-rater reliability in ultrasound body composition measurements and investigating the differences between malnourished and non-malnourished infants. Sonographic images for measurements of fat and muscle thickness were compared between 9 malnourished and 9 non-malnourished hospitalized infants. The mean of fat and muscle thickness sums were 12.44 ± 7.58 mm and 28.98 ± 7.18 mm, respectively. The intra- and inter-rater intraclass correlation coefficient were above 0.9 for both measurements, indicating high intra- and inter-rater reliability. Compared to non-malnourished infants, malnourished infants have 45% of fat thickness sum and 71% of muscle thickness sum. Ultrasound measurements of body composition in infants were different between hospitalized malnourished and non-malnourished infants. This approach has the potential to be utilized more broadly, from assessing the nutritional status of critically ill infants in intensive care units to screening for malnutrition in high-risk infant populations.
Keyphrases
- body composition
- optical coherence tomography
- magnetic resonance imaging
- adipose tissue
- bone mineral density
- skeletal muscle
- intensive care unit
- randomized controlled trial
- fatty acid
- deep learning
- body mass index
- magnetic resonance
- risk assessment
- convolutional neural network
- weight gain
- extracorporeal membrane oxygenation
- human health